Over the past decade, the NCI supported Physicians' Health Study (PHS) of biomarkers and prostate cancer has prospectively evaluated and demonstrated the importance of several risk factors or markers including prostate specific antigen, sex hormone levels, the androgen receptor polymorphism, and insulin-like growth factor-I and its binding protein-3. The importance of sex hormones in prostate cancer is well established, but several critical issues remain to be addressed. These include the impact of short- vs. long-term exposure, the validity of a single vs. multiple measurements, identification of genetic and lifestyle influences on hormone levels, effects of changes in hormone levels, and the association of hormones with specific clinical and pathological characteristics. In this competing renewal grant, we propose to address these specific issues using long-term archived blood samples collected at baseline (1982-1984) in the PHS I and a second blood sample collected from over 60% of the original participants during 1995-1998 in the PHS II (an ongoing randomized trial of nutritional supplements). We will also assess the role of several newly identified polymorphisms in the sex hormone metabolism pathways, receptors, and a prostate cancer susceptibility gene. Recent data suggest that transforming growth factor (TGF)-Beta1 and interleukin (IL)-6 play important roles in prostate tumorigenesis, angiogenesis. and progression. We propose to prospectively assess the role of prediagnostic levels of these factors and several of their genetic polymorphisms. We will further explore the clinical significance of circulating levels (and genetic variations) of hormones, TGF-Beta1 and IL-6 in relation to their expression in tumor tissue, angiogenesis and prostate cancer progression. In the PHS 1, we have already assembled an extensive prospective clinical and serological database on prostate cancer with almost two decades of follow-up of 14,916 men. The PHS II collected and archived 7.241 blood samples. The two blood samples collected between the PHS I and the PHS It among the same 9.436 participants provides a unique and cost-effective opportunity to study the hypotheses proposed here. The large number of cases (n=1 ,765) will provide sufficient statistical power to provide firm answers to these questions.